Course automatic recommendation management system and method based on artificial intelligence

Abstract

The invention discloses an automatic curriculum recommendation management system and method based on artificial intelligence, and belongs to the technical field of artificial intelligence. The system comprises an information marking module, an edge database module, a time period analysis module, an intelligent flow pushing module and an error feedback module. The information marking module collects user information and user browsing information records; the edge database module stores collected user information and browsing information records; the time period analysis module analyzes the use condition of the course app, the information browsing type and the browsing habit of the user; the intelligent flow pushing module autonomously and dynamically adjusts a course app information pushing mode according to course app using habits and information accepting habits of a user in different time periods; the error feedback module periodically analyzes the error condition of the information pushed to the user and feeds back the error condition; according to the method and the device, the information content is intelligently pushed according to the habits of the user in the course app using process of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and specifically to a course automatic recommendation management system and method based on artificial intelligence. Background Technique

[0002] The automatic recommendation of artificial intelligence courses refers to using artificial intelligence technology to analyze data such as user behavior and interests, automatically generating personalized recommendation plans, and achieving efficient and accurate course matching; At present, following the development of people's course needs, Internet companies attach particular importance to realizing real-time online and offline interaction of courses and instant push of new information; this is an opportunity in the new era but also brings a lot of trouble. Due to the large openness of instant course pictures and texts, course short videos, and real-time course live broadcasts, the number of users participating in them is huge, and the amount of information brought by a large number of users is extremely "explosive". If such a huge amount of information is not screened and targeted to users, the instant information and real-time interaction information received by each user on the course app will be relatively messy; therefore, there is an urgent need for a method to dynamically adjust the information push of the course app according to the information required by users at different stages. Summary of the Invention

[0003] The purpose of the present invention is to provide a course automatic recommendation management system and method based on artificial intelligence to solve the problems raised in the above background technique.

[0004] To solve the above technical problems, the present invention provides the following technical solutions: A course automatic recommendation management system based on artificial intelligence, the course automatic recommendation management system based on artificial intelligence includes an information marking module, an edge database module, a time period analysis module, an intelligent push module, and an error feedback module; The information marking module is used to collect user information and user browsing information records; the edge database module is used to store the collected user information and browsing information records in the user edge database; the time period analysis module is used to analyze different time periods when the user uses the course app, the information content accepted by the user during the corresponding time periods, and browsing habits; the intelligent push module automatically and dynamically adjusts the course app information push mode according to the course app usage habits and information acceptance habits of users at different time periods; the error feedback module periodically analyzes the error situation of the information pushed to the user and feeds back the error situation; The information marking module is connected to the edge database module; the edge database module is connected to the time period analysis module; the time period analysis module is connected to the intelligent push module; the intelligent push module is connected to the error feedback module.

[0005] The information marking module includes a user information collection unit and a user browsing information recording unit; the user information collection unit is used to collect real-name information of the user and conduct authenticity verification when the user first uses the course app; the user browsing information recording unit is used to record the information type, the browsing duration of a single piece of information, the total duration of a single piece of information, the duration of the user's single use of the course app, and the time period when the user uses the course app when the user browses information using the course app; wherein the user browsing information includes graphic information, video information, and live broadcast information.

[0006] The edge database module includes a user information repository and a user habit record repository; the user information repository is used to store the identity information collected by the course app users; the user habit record repository is used to store the content of the information browsed by the user when receiving information using the course app and the user's personal habits and browsing operations when browsing information.

[0007] The time period analysis module includes a time period usage habit analysis unit and a multi-time period information reception analysis unit; the time period usage habit analysis unit analyzes the usage situation of the course app by the user in different time periods, including analyzing the fragmentation degree of the time when the user uses the course app and judging the concentration degree of the user using the course app; the multi-time period information reception analysis unit analyzes the information reception requirements of the user in different time periods according to the type and duration of the information received by the user in different time periods.

[0008] The intelligent push flow module includes a multi-time period information push flow unit and a predicted information judgment unit; the multi-time period information push flow unit pushes the information content that the user is concerned about to the user according to the information reception type and browsing habit of the user in different time periods; the predicted information judgment unit judges the push situation of the first a pieces of information according to the concentration degree of the user's current browsing information, where a is a constant.

[0009] The error feedback module includes a push flow error analysis unit and a dynamic correction unit; the push flow error analysis unit judges whether the user is concerned about the single piece of information according to the browsing duration and operation of the information intelligently pushed by the course app by the user, calculates the error of the intelligent push flow of the course app by statistically analyzing the occupancy of the information that the user is not concerned about in different time periods; the dynamic correction unit screens out the information that the user is not concerned about in different time periods in a similar way, and associates the screening record with the user information in the edge database; stores the screening record in the database and updates the screening record periodically, and the update method is to overwrite the old with the new.

[0010] A course automatic recommendation management method based on artificial intelligence, the method includes the following steps: S100. For users who use the course app for the first time, request authorization from the users for real-name information and browsing data collection; after the users' authorization is obtained, collect the users' identity information and the habits of receiving information when using the course app, and store the collected data in a database; S200. Analyze the types of information browsed by users at different times and their browsing habits, and intelligently push information services to users according to the analysis results; S300. Conduct error analysis on the pushed information services, and perform dynamic correction operations in combination with the results of the error analysis.

[0011] The specific steps of collecting the identity information of users who use the course app for the first time and the habits of receiving information when using the course app in S100 and storing the collected data in a database are as follows: S101. When a user first downloads and uses the course app, through a pop-up message, request authorization from the user for real-name information and authorization for browsing information records when using the course app. After obtaining the user's authorization, collect the user's identity information, including the user's name, gender, ID number, and contact information; authenticate the collected user identity information through the background; if it is correct, the user information collection is successful; otherwise, collect it again until the authentication is successful; store the collected user identity information in an edge database near the user side in the network; storing the data in the edge database is convenient for retrieval and storage, and can achieve low latency; S102. After collecting and successfully authenticating the user's identity information, record and learn the types of the user's browsing information and browsing operations at a period T, and store the time period when the user uses the course app, the duration of each use, and the types of information browsed by the user when using the course app in an edge database near the user side.

[0012] The specific steps of analyzing the types of information browsed by users at different times and their browsing habits in S200 and intelligently pushing information services to users according to the analysis results are as follows: S201. Retrieve the usage situation of the course app by the user throughout the day, the types c of information browsed by the user at different times, the duration t of the user browsing a single piece of information, and the duration of a single piece of information collected in the database ; Use the K-means clustering algorithm to classify the usage situation of the course app by the user in multiple time periods throughout the day within the period T. Based on the classification results, divide the time periods when the user uses the course app throughout the day into concentrated information reception periods and scattered information reception periods; calculate the average duration of the user using the course app during the concentrated information reception period respectively according to the statistical data and the average duration of the user using the course app during the scattered information reception period ; S202. Through the formula Calculate the concentration of a single user for a single piece of information; the course app uses big data to implant the core information points of different types of single pieces of information through a formula Calculate the proportion of core information in a single piece of information; according to the formula Calculate the required browsing duration of a single piece of information at different browsing magnification ratios ; where v is the browsing magnification ratio used by the user when browsing the information; for the duration t of the user browsing a single piece of information and the single-magnification browsing duration of the single piece of information and the multi-magnification browsing duration perform a size comparison; where previously is the information paving duration, and this paving duration is the information content duration from the start of the information to the core information point, afterwards is the duration of the explanation around the core information, then is the core information content duration of a single piece of information; When the user's browsing speed is single-speed, if , then the user has one or more complete browsing behaviors for the current information, and at this time the user's concentration , then the user is interested in the current information; if , then calculate the user's concentration q for the current information and the proportion w of the core information in the current information; if , then the user is not interested in the current information; if , then the user has a behavior of dragging the progress bar for the current information; when the user's browsing concentration for a single piece of information is less than 1 but greater than the proportion of the core information of the information, it means that the current user is interested in the current information but only interested in the core information in the information, so the user will perform a behavior of dragging the progress bar; When the user's browsing speed is v, calculate the required browsing duration of the information at the current magnification ; if , then the user has one or more complete browsing behaviors for the current information, and at this time the user's concentration , then the user is interested in the current information; if , calculate the user's concentration for the current information at this time through the formula ; where represents the concentration of the user browsing the information at the multi-magnification browsing rate; calculate the proportion of the core information in a single piece of information at the multi-magnification browsing rate through the formula ; if , then the user is not interested in the current information; if , then the user has a behavior of dragging the progress bar for the current information; S203. Based on the user's reception concentration of the information, and screen out the information of and ; where c is the set of information types in the concentrated information reception period, is the set of information types in the scattered information reception period, ; Taking the number of the same type of information as the x-axis and the user's concentration q as the y-axis, obtain the concentration attenuation relationship curve of the user's continuous reception of the same type of information through a non-linear regression model; its curve expression is , where is the regression coefficient; calculate the value of b corresponding to on the curve, and the integer value of b is the number a of pre-pushed information of the course app; take on the curve, and the corresponding value, and the integer value of is the maximum threshold z of the number of consecutive receptions of the same type of information by the user; then when the course app performs the information pre-push service, push the information content of interest to the user at different times for the user, the number of pre-pushed information is a, and dynamically adjust the pre-pushed information to maintain the number of the same type of information ; when the user's concentration attenuation to 0 for the continuous reception of the same type of information, it means that the number of information receptions at this time is the limit of the current user, and the course app can facilitate the information regulation with this value as the number of pre-pushed information; at the same time, in order to ensure that the same type of information does not continuously appear when the user browses the information, resulting in a decrease in the user's interest, limit the number of the same type of information in the pre-pushed information of the course app.

[0013] The specific steps for error analysis of the push information service in S300 and dynamic correction operations in combination with the error analysis results are as follows: S301. Taking the number a of pre-pushed information of the course app as a group, randomly query the user browsing situation of m groups of information in the current cycle T of the user; calculate the user concentration q corresponding to the respective browsing magnification; judge the time period when the user browses m groups of information, and respectively count the number of information in the user's concentrated information reception period and scattered information reception period and ; calculate the error rate of the push information service of the course app through the formula ; calculate the error rate of the currently randomly selected information group based on the sum of the product of the proportion of the error information in the corresponding group and the proportion of the error in the corresponding group in the time period, and use this value to represent the error rate of the entire system; where, is the error proportion coefficient in the concentrated information reception period, is the error ratio coefficient for the scattered information reception period; its calculation formulas are respectively , ; and ; Since during the concentrated information reception period, users spend a longer time using the course app, it is necessary to ensure the freshness of information reception for users, so it is more important to reduce the occurrence of error information. Therefore, the error ratio for this period should be large; when , mark the error information and perform screening of the same type; when , clean the information content pre-pushed to the user, and recalculate and push the information according to the user's preferred type of information reception and browsing habits within the current cycle; where is a constant set by the system; S302. Record the screened information in the database and update the screening record with a period of T. The update method is to overwrite the old data with new data; since within different cycles of the user, the user's focus on different information changes with the types and quantities of the information they browse, so screen and record the information that the user is not interested in within different cycles for a short time, and in the next cycle, overwrite the old record with a new screening record to ensure the diversification of the types of information browsed by the user through cycling.

[0014] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The present invention realizes the classification of the scattered degree of the period when the user uses the course app through the combined action of multiple modules, and analyzes the user's focus on information within different periods; classifies the types and contents of information that the user is interested in within different periods by calculating the values of the user's focus on different information; calculates the quantity of information pre-pushed by the course app background and the push quantity limit of the same type of information by analyzing the change degree of the user's focus on the same type of information; calculates the error of the information pushed by the course app in the current cycle by randomly extracting the historical browsing data in the user's usage cycle, and dynamically adjusts the information push for the user by comparing the calculation result with the threshold; the present invention combines a multi-functional module with an analysis method to intelligently push information content according to the user's habits during the user's use of the course app. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention. In the drawings: Figure 1 is a schematic structural diagram of a course automatic recommendation management system based on artificial intelligence according to the present invention; Figure 2 is a flowchart of the steps of a course automatic recommendation management method based on artificial intelligence according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0017] Please refer to Figure 1 - Figure 2 , the present invention provides a technical solution: An artificial intelligence-based automated course recommendation management system, the artificial intelligence-based automated course recommendation management system includes an information marking module, an edge database module, a time period analysis module, an intelligent push module, and an error feedback module; The information marking module is used to collect user information and user browsing information records; the edge database module is used to store the collected user information and browsing information records in the user edge database; the time period analysis module is used to analyze different time periods when the user uses the course app, the information content accepted by the user in the corresponding time period, and the browsing habits; the intelligent push module automatically adjusts the information push mode of the course app according to the user's course app usage habits and information acceptance habits at different time periods; the error feedback module periodically analyzes the error situation of the information pushed to the user and feeds back the error situation; The information marking module is connected to the edge database module; the edge database module is connected to the time period analysis module; the time period analysis module is connected to the intelligent push module; the intelligent push module is connected to the error feedback module.

[0018] The information marking module includes a user information collection unit and a user browsing information recording unit; the user information collection unit is used to collect the real-name information of the user and authenticate its authenticity when the user first uses the course app; the user browsing information recording unit is used to record the information type, single-piece information browsing duration, single-piece information total duration, user's single-time course app usage duration, and the time period when the user uses the course app when the user browses information using the course app; wherein the user browsing information includes graphic information, video information, and live broadcast information.

[0019] The edge database module includes a user information repository and a user habit record repository; the user information repository is used to store the identity information collected by the course app users; the user habit record repository is used to store the content of the information browsed by the user when receiving information using the course app and the user's individual habits and browsing operations when browsing information.

[0020] The time period analysis module includes a time period usage habit analysis unit and a multi-time period information reception analysis unit; the time period usage habit analysis unit analyzes the usage of the course app by the user in different time periods, including analyzing the fragmentation degree of the time when the user uses the course app and judging the concentration degree of the user using the course app; the multi-time period information reception analysis unit analyzes the information reception requirements of the user in different time periods according to the type and duration of the information received by the user in different time periods.

[0021] The intelligent push flow module includes a multi-time period information push flow unit and a prediction information judgment unit; the multi-time period information push flow unit pushes the information content that the user is concerned about to the user according to the information reception type and browsing habit of the user in different time periods; the prediction information judgment unit judges the push situation of the first a pieces of information according to the concentration degree of the user's current browsing information, where a is a constant.

[0022] The error feedback module includes a push flow error analysis unit and a dynamic correction unit; the push flow error analysis unit judges whether the user is concerned about a single piece of information according to the browsing duration and operations of the information intelligently pushed by the course app by the user, calculates the error of the intelligent push flow of the course app by statistically analyzing the occupancy degree of the information that the user is not concerned about in different time periods; the dynamic correction unit screens out the information that the user is not concerned about in different time periods in a similar way, and associates the screening record with the user information in the edge database; stores the screening record in the database and updates the screening record periodically, and the update method is to cover the old with the new.

[0023] A course automatic recommendation management method based on artificial intelligence, the method includes the following steps: S100. For a user who uses the course app for the first time, apply for authorization for real-name information and browsing data collection to the user; after the user authorizes and permits, collect the user's identity information and the habit of receiving information when the user uses the course app, and store the collected data in the database; S200. Analyze the information type and browsing habit browsed by the user in different time periods, and intelligently push information services to the user according to the analysis results; S300. Analyze the error of the pushed information service, and perform dynamic correction operations in combination with the error analysis results.

[0024] The specific steps of collecting the user's identity information and the habit of receiving information when the user uses the course app for the first time in S100 and storing the collected data in the database are as follows: S101. When a user first downloads and uses the course app, a pop-up message is used to request authorization for the user's real-name information and browsing information record of using the course app. After obtaining the user's authorization permission, the user's identity information is collected, including the user's name, gender, ID number, and contact information; the collected user identity information is authenticated for authenticity through the background; if it is correct, the user information collection is successful; otherwise, it is collected again until the authentication is successful; the collected user identity information is stored in an edge database near the user side in the network; storing the data in the edge database facilitates retrieval and storage, and low latency can be achieved. S102. After the user identity information is collected and authenticated successfully, the types and browsing operations of the user's browsing information are recorded and learned at a period T, and the time period of the user using the course app, the duration of each use, and the types of information browsed by the user using the course app are stored in an edge database near the user side.

[0025] The specific steps of analyzing the information types and browsing habits of the user at different time periods in S200 and intelligently pushing information services to the user according to the analysis results are as follows: S201. Retrieve the usage situation of the user's full-day course app in the database, the types c of information browsed by the user at different time periods, the duration t of the user browsing a single piece of information, and the duration of a single piece of information ; Use the K-means clustering algorithm to classify the usage situation of the user's full-day course app in multiple time periods within the period T. Based on the classification results, divide the user's full-day course app usage time periods into concentrated information reception time periods and scattered information reception time periods; calculate the average duration of the user using the course app in the concentrated information reception time period according to the statistical data and the average duration of the user using the course app in the scattered information reception time period ; S202. Calculate the concentration degree of a single piece of information for the user through the formula ; The course app uses big data to implant the core information points of different types of single pieces of information ; Calculate the proportion degree of the core information in a single piece of information through the formula ; Calculate the browsing duration required for a single piece of information at different browsing magnifications according to the formula ; where v is the browsing magnification used by the user when browsing information; compare the duration t of the user browsing a single piece of information with the single-magnification browsing duration and the multi-magnification browsing duration of a single piece of information; among them, is the information paving duration before, and this paving duration is the information content duration from the beginning of the information to the core information point, and Subsequently, for the duration of the explanation around the core information, is the duration of the core information content of a single piece of information; When the user's browsing speed is normal, if , the user has one or more complete browsing behaviors for the current information. At this time, the user's concentration , the user is interested in the current information; if , calculate the user's concentration q on the current information and the proportion w of the core information in the current information; if , the user is not interested in the current information; if , the user has a behavior of dragging the progress bar for the current information; when the user's browsing concentration on a single piece of information is less than 1 but greater than the proportion of the core information of the information, it means that the current user is interested in the current information but only interested in the core information in the information. Therefore, the user will perform the behavior of dragging the progress bar; When the user's browsing speed is v, calculate the required browsing duration of the information at the current magnification ; if , the user has one or more complete browsing behaviors for the current information. At this time, the user's concentration , the user is interested in the current information; if , through the formula calculate the user's concentration on the current information at this time; among them, represents the user's browsing concentration on the information at the multi-magnification browsing rate; through the formula calculate the proportion of the core information in a single piece of information at the multi-magnification browsing rate; if , the user is not interested in the current information; if , the user has a behavior of dragging the progress bar for the current information; S203. Based on the user's reception concentration of the information, and the information is screened out, and the set of information types that the user browses with interest in different time periods is counted as and ; where c is the set of information types in the concentrated information reception period, is the set of information types in the scattered information reception period, ; Taking the number of the same type of information as the x-axis and the user's concentration q as the y-axis, obtain the concentration decay relationship curve of the user's continuous reception of the same type of information through a non-linear regression model; its curve expression is , where is the regression coefficient; calculate on the curve The value of b corresponding to the time, and the rounded value of b is the number a of pre-pushed messages of the course app; take at the curve the corresponding value, for rounding, the value is the maximum threshold z of the number of consecutive same-type message receptions by the user; then when the course app performs the information pre-push service, for the user, push the information content of interest to the user at different time periods, the number of pre-pushed messages is a, and dynamically adjust the pre-pushed messages to maintain the number of same-type messages ; when the user's concentration attenuation for continuously receiving the same-type message reaches 0, it means that the number of received messages at this time is the limit of the current user. Then, the course app can use this value as the number of pre-pushed messages to facilitate the regulation of information; at the same time, in order to ensure that the same-type messages do not continuously appear when the user browses the information, resulting in a decrease in the user's interest, limit the number of same-type messages in the pre-pushed messages of the course app.

[0026] In S300, perform error analysis on the push message service, and the specific steps for dynamic correction operation in combination with the error analysis results are as follows: S301. Take the number a of pre-pushed messages of the course app as a group, and randomly query the user browsing situation of m groups of messages in the current cycle T of the user; calculate the user concentration q corresponding to the corresponding browsing magnification respectively; judge the time period when the user browses m groups of messages, and respectively count the number of messages in the user's concentrated information reception period and scattered information reception period and ; through the formula calculate the error rate of the push message service of the course app; calculate the error rate of the currently randomly selected information group based on the sum of the product of the proportion of error information in the corresponding group and the proportion of time period error in the corresponding group, and use this value to represent the error rate of the entire system; among them, is the error proportion coefficient of the concentrated information reception period, is the error proportion coefficient of the scattered information reception period; their calculation formulas are respectively , , ; and ; since in the concentrated information reception period, the user uses the course app for a long time, it is necessary to ensure the freshness of the user's information reception, so it is more important to reduce the appearance of error information. Therefore, the error proportion in this period should be large; when , mark the error information and perform same-type screening; when , clean the information content pre-pushed to the user, and recalculate and push the information according to the user's information reception preference type and browsing habit in the current cycle; among them is a constant set by the system; S302. Record the screened information in the database and update the screening records at a period T. The update method is to overwrite the old data with new data. Since the user's focus on different information changes with the types and quantities of the information they browse in different periods, the information that the user is not interested in during different periods is briefly screened and recorded, and the old records are overwritten with new screening records in the next period, and the cycle is repeated to ensure the diversification of the types of information browsed by the user.

[0027] In the embodiment: If there is a short video course app device with a platform intelligent push system based on big data, and a certain user uses the course app for the first time, the identity information of the current user is collected and authenticated, and the real identity information is stored in the edge database. After the user is authenticated, the information content and habits browsed by the current user are recorded after a period of 1 day. Among them, 7-8 am and 6-8 pm are the concentrated information reception periods, and 9-11 am, 2-5 pm, and 9-10 pm are the scattered information reception periods. Among them, the average usage duration of the course app by the user during the concentrated information reception period is 1h, and the average usage duration of the course app during the scattered information reception period is 1 / 4h. Statistics show that the types of information browsed by the user during the concentrated information reception period are , and the types of information browsed during the scattered information reception period are ; The core information points are implanted into the information content through big data. Among them, the information duration during the concentrated information reception period is , the browsing duration of the user is , and the core information points of the information are ; The information duration during the scattered information reception period is , the browsing duration of the user is ; The core information points of the information are ; The current user browses the information at a single rate. According to the formula Calculate the user's concentration during the concentrated information reception period , calculate the user's concentration during the scattered information reception period ; According to the formula Calculate the proportion of the core information of the information during the concentrated information reception period , calculate the proportion of the core information of the information during the scattered information reception period ; By comparing the user's concentration and the proportion of the core information, the set of information types that the user is interested in during the concentrated information reception period is , and the set of information types that the user is interested in during the scattered information reception period is ; By non-linearly regressing historical data, with the number b of the same type of information as the x-axis and the user's concentration q as the y-axis, obtain the concentration attenuation relationship curve of the user continuously receiving the same type of information, and the curve expression is ; When corresponding to , then the number of pre-pushed messages of the course app is ; When q takes and respectively, then the values of u are and , then the maximum threshold for pre-pushing each type of information during the concentrated information reception period of the course app is , and the maximum threshold for pre-pushing each type of information during the scattered information reception period is ; Taking 4 messages as a group, randomly select 4 groups of the user's browsing situation in the current day; 2 groups come from the concentrated information reception period and 2 groups come from the scattered information reception period; respectively count the number of messages, the number of messages that the user is not interested in during the concentrated information reception period is 1, and the number of messages that the user is not interested in during the scattered information reception period is 1; According to the formula calculate the error proportion coefficient during the concentrated information reception period; According to the formula calculate the error proportion coefficient during the scattered information reception period; According to the formula calculate the error rate of the course app's push message service. Since , then clean the information content pre-pushed to the user, and recalculate and push the information according to the user's information reception preference type and browsing habit in the current cycle.

[0028] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0029] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An artificial intelligence-based course automatic recommendation management system, characterized by: The course automatic recommendation management system based on artificial intelligence includes an information marking module, an edge database module, a time period analysis module, an intelligent streaming module and an error feedback module; The information marking module is used to collect user information and user browsing information records; the edge database module is used to store the collected user information and browsing information records in the user edge database; The time period analysis module is used to analyze the different time periods when users use the course app, the information content and browsing habits accepted by users in the corresponding time periods; the intelligent streaming module autonomously and dynamically adjusts the course app information push mode according to the course app usage habits and information acceptance habits of users in different time periods; the error feedback module periodically analyzes the error situation of the information pushed to the user and provides feedback on the error situation; The information marking module is connected to the edge database module; the edge database module is connected to the time period analysis module; The time period analysis module is connected to the intelligent streaming module; and the intelligent streaming module is connected to the error feedback module.

2. According to claim 1, the course automatic recommendation management system based on artificial intelligence is characterized by: The information marking module includes a user information collection unit and a user browsing information recording unit; the user information collection unit is used to collect the user's real-name information and perform authenticity authentication when the user uses the course app for the first time; the user browsing information recording unit is used to record the browsed information type, the browsing time of a single piece of information, the total browsing time of a single piece of information, the time the user uses the course app for a single time, and the time period when the user uses the course app when browsing information using the course app; wherein the user browsing information includes graphic information, video information, and live broadcast information.

3. The course automatic recommendation management system based on artificial intelligence according to claim 2 is characterized by: The edge database module includes a user information repository and a user habit record library; the user information repository is used to store identity information collected by course app users; The user habit record library is used to store the content of information browsed by the user when receiving information using the course app and the user's personal habits and browsing operations in browsing information.

4. The course automatic recommendation management system based on artificial intelligence according to claim 3 is characterized by: The time period analysis module includes a time period usage habit analysis unit and a multi-time period information reception analysis unit; The time period usage habit analysis unit analyzes the usage of the course app according to the user in different time periods, including analyzing the degree of fragmentation of the user's time in using the course app and judging the user's concentration in using the course app; the multi-time period information reception analysis unit analyzes the user's information reception needs in different time periods according to the type and duration of information received by the user in different time periods.

5. The course automatic recommendation management system based on artificial intelligence according to claim 4 is characterized by: The intelligent streaming module includes a multi-period information streaming unit and a predicted information judgment unit; the multi-period information streaming unit pushes information content that the user is interested in according to the information receiving type and browsing habits of the user in different time periods; the predicted information judgment unit judges the push status of the next a pieces of information according to the user's current concentration on browsing information, where a is a constant.

6. The course automatic recommendation management system based on artificial intelligence according to claim 5 is characterized by: The error feedback module includes a streaming error analysis unit and a dynamic correction unit; the streaming error analysis unit determines whether the user pays attention to the single piece of information according to the browsing time and operation of the information of the course app intelligent streaming, and calculates the error of the course app intelligent streaming by counting the proportion of information that the user does not pay attention to in different time periods; The dynamic correction unit performs similar screening for information that the user does not pay attention to in different time periods, and records the screening in the edge database and links it with the user information; the screening records are stored in the database, and the screening records are updated periodically, and the update method is to overwrite the old with the new.

7. An artificial intelligence-based automatic course recommendation management method, characterized in that: The method comprises the following steps: S100: For a user who uses the course app for the first time, apply for authorization from the user to collect real-name information and browsing data; after the user authorizes and permits, collect the user's identity information and the user's habit of receiving information when using the course app, and store the collected data in a database; S200, analyzing the information types and browsing habits of users at different time periods, and intelligently pushing information services to users based on the analysis results; S300: Perform error analysis on the push information service, and perform dynamic correction operations based on the error analysis results.

8. The method for automatic course recommendation management based on artificial intelligence according to claim 7, characterized in that: The specific steps of collecting the identity information of the user who uses the course app for the first time and the user's habit of receiving information using the course app and storing the collected data in the database in S100 are as follows: S101. When a user downloads and uses the course app for the first time, a pop-up window is used to request the user's real-name information and authorization for the user's browsing information record when using the course app. After obtaining the user's authorization, the user's identity information is collected, including the user's name, gender, ID number, and contact information. The collected user identity information is authenticated through the background; if it is correct, the user information is collected successfully; Otherwise, re-collect until the authentication is successful; store the collected user identity information in the edge database close to the user end in the network; S102. After the user identity information is collected and authenticated successfully, the types of information browsed and browsing operations of the user are recorded and learned in period T, and the time period of the user using the course app, the duration of a single use, and the type of information browsed by the user using the course app are stored in an edge database close to the user end.

9. The method for automatic course recommendation management based on artificial intelligence according to claim 8, characterized in that: The specific steps of analyzing the information types and browsing habits of users at different time periods and intelligently pushing information services to users according to the analysis results in S200 are as follows: S201, retrieve the user's full-day course app usage collected in the database, the type of information browsed by users in different time periods c, the time t for users to browse a single piece of information and the time of a single piece of information ; K-means clustering algorithm is used to classify the course app usage of users in multiple time periods throughout the day within period T. Based on the classification results, the user's full-day course app usage period is divided into concentrated information reception period and scattered information reception period; according to statistical data, the average duration of users using the course app during the concentrated information reception period is calculated The average time users spend using the course app during the period of sporadic information reception ; S202, through the formula Calculate the user's concentration on a single piece of information; the course app uses big data to analyze the core information points of different types of single information. implant; through formula Calculate the proportion of core information in a single piece of information; According to the formula Calculate the browsing time required for a single piece of information at different browsing magnifications ; where v is the browsing magnification used by the user when browsing information; the time t for users to browse a single piece of information and the single-multiple browsing time for a single piece of information and multiple-rate browsing time Perform size comparison; among them, The time it takes to lay the groundwork for the information is long. After that, the length of time to explain the core information is as follows: The length of the core information content of a single message; When the user browses at a single speed, , the user has browsed the current information once or more completely, and the user's concentration is , the user is interested in the current information; if , then calculate the user's concentration on the current information q and the proportion of core information in the current information w; if , the user is not interested in the current information; if , the user drags the progress bar of the current information; When the user browses at a speed of v, calculate the browsing time required for the information at the current speed ;like , the user has browsed the current information once or more completely, and the user's concentration is , the user is interested in the current information; if , through the formula Calculate the user's concentration on the current information at this time; where, It is expressed as the user's concentration when browsing information at multiple browsing rates; Calculate the proportion of core information in a single piece of information at multiple browsing rates; if , the user is not interested in the current information; if , the user drags the progress bar of the current information; S203, based on the user's concentration on receiving the information, and The information is filtered out, and the information types that users browse in different time periods are counted as follows: and ; Where c is the information type set of the centralized information reception period, It is a collection of information types during the period of receiving scattered information. ; The number of the same type of information With q as the x-axis and the user's concentration q as the y-axis, the concentration decay relationship curve of users who continuously receive the same type of information is obtained through the nonlinear regression model; the curve expression is: ,in is the regression coefficient; The value of b is corresponding to the time, and the value after rounding b is the number of information pre-pushed by the course app a; take The corresponding The value of b is rounded to the integer value, which is the maximum threshold z for the number of consecutive messages of the same type received by the user. When the course app performs information pre-push service, it pushes information of interest to the user at different time periods. The number of pre-push messages is a, and the pre-push messages are dynamically adjusted to maintain the number of messages of the same type. .

10. The method for automatic course recommendation management based on artificial intelligence according to claim 9, characterized in that: The specific steps of performing error analysis on the push information service in S300 and performing dynamic correction operation based on the error analysis result are as follows: S301, taking the number of information a pre-pushed by the course app as a group, randomly query the user's browsing status of m groups of information in the current period T; calculate the user's concentration q under the corresponding browsing magnification; determine the time period when the user browses the m groups of information, and count the time period when the user is in concentrated information reception and scattered information reception. Number of messages and ; Through the formula Calculate the error rate of the course app push information service; where, is the error proportional coefficient of the centralized information reception period, is the error ratio coefficient of the scattered information reception period; its calculation formula is , ; and ;when When , the error information is marked and the same type is screened out; when When the user's pre-pushed information is cleaned, the information is recalculated and pushed according to the user's information receiving preference type and browsing habits in the current period; Set constants for the system; S302: Record the screening information in a database and update the screening records periodically at T, by overwriting old data with new data.

Images